In the early-mid 80's most of the Internet work force either had PhDs or working toward their PhDs. The Internet was largely run by faculty, their research groups (i.e. grad students), university & contract R&D staff (e.g. BBN and staff working at universities). When the Internet got more popular and universities/others started teaching courses and programs for network management - we basically got the degrees that are required by ISPs. So, we basically went from PhDs --> undergraduate degrees.
There were rarely people with training based on tinkering in their garage running the routers/links/etc. of the Internet:-)
The above was simply a trend as a result of the maturation process of the technology. In the 80's, very few parts of the Internet worked well or were easy to fix/understand. As things started working better and changed less often, we could start teaching classes and have companies with support-systems targeting undergrads rather than PhD students.
Another important factor in all this, is that no one wants to hire you with no training/experience because you will do a terrible job. You are basically paying the university to get experience quickly. You could get that experience working in your garage and reading books - but this is slow. In addition, a hiring company has no easy way to judge whether you really read the books and tinkered on the right things. A university is also a certification authority - they verify that you did the work. This is why companies require a degree and sometimes overlook it for candidates with important experience.
This is not the same type of attack -- the AS7007 problem was a route hijack attack.
The sigcomm paper describes a more basic route convergence issue with path vector protocols
The paper describes the use of packet loss to create a BGP session failure and the impact of repeated announce/withdraw traffic to slow other routers. This is also not new. However, the appropriate point of reference is "RFC 1266 - Experience with the BGP Protocol" (http://www.faqs.org/rfcs/rfc1266.html). Read section 9 -- this points to how packet loss results in BGP failures and points to how ensuring BGP packets have priority fixes this. This was published in 1991:-) and is generally well known.
Similarly, I haven't read the referenced NDSS paper (http://www-users.cs.umn.edu/~hopper/lci-ndss.pdf) but I am also surprised that BGP holddown timers don't prevent some of the related route churn problems.
1) Blaming corporations is the wrong path. Corporations are forced to compete on a global playing field. Our world has evolved to the point that having large research labs (e.g., IBM TJ Watson, AT&T Research, Bell Labs, MSR) is not an effective competitive advantage. The problem is that, in many fields of research, it is too easy to play catch-up. These research labs put out a few great ideas that are brought to market (but the labs cost a large number of $). These ideas are copied (i.e. people come up with similar techniques that circumvent patents) sufficiently well in a few months. As a result, the profit gained by being first to market is limited and the business of creating ideas is no longer profitable for individual companies:-(
Unfortunately, we still need research to advance and grow entire markets - the problem is identifying a way to fund or compensate companies/individuals/organizations for their research contributions without relying on market profits.
Basically, the world changed - rapid prototyping and easy access to ideas/research are common to most topic areas. It's time to face the facts and come up with something new that works.
2) Foreigners and graduate school. This is a huge problem. The American graduate education system has survived on the influx of students from countries like India and China. Unfortunately, the combination of improving opportunities in these countries and a negative immigration environment in the US has made this transfer of talent slow down. In 20 years, I am not convinced that the best and brightest will continue to come to the U.S.
3) No jobs for PhDs. I agree that jobs are getting harder to find for PhDs. Given comment (1) above, I am not sure there are going to be more PhD research jobs in the future. If we solve (1), there will be more jobs.
Seshadri, Arvind, Mark Luk, Elaine Shi, Adrian Perrig, Leendert van Doorn, and Pradeep Khosla. "Pioneer: Verifying Integrity and Guaranteeing Execution of Code on Legacy Platforms." In Proceedings of the ACM Symposium on Operating Systems Principles (SOSP), Brighton, United Kingdom, October 2005.
Slashdot Mirror
I hate to burst your bubble....
In the early-mid 80's most of the Internet work force either had PhDs or working toward their PhDs. The Internet was largely run by faculty, their research groups (i.e. grad students), university & contract R&D staff (e.g. BBN and staff working at universities). When the Internet got more popular and universities/others started teaching courses and programs for network management - we basically got the degrees that are required by ISPs. So, we basically went from PhDs --> undergraduate degrees.
There were rarely people with training based on tinkering in their garage running the routers/links/etc. of the Internet :-)
The above was simply a trend as a result of the maturation process of the technology. In the 80's, very few parts of the Internet worked well or were easy to fix/understand. As things started working better and changed less often, we could start teaching classes and have companies with support-systems targeting undergrads rather than PhD students.
Another important factor in all this, is that no one wants to hire you with no training/experience because you will do a terrible job. You are basically paying the university to get experience quickly. You could get that experience working in your garage and reading books - but this is slow. In addition, a hiring company has no easy way to judge whether you really read the books and tinkered on the right things. A university is also a certification authority - they verify that you did the work. This is why companies require a degree and sometimes overlook it for candidates with important experience.
This is not the same type of attack -- the AS7007 problem was a route hijack attack.
The sigcomm paper describes a more basic route convergence issue with path vector protocols
The paper describes the use of packet loss to create a BGP session failure and the impact of repeated announce/withdraw traffic to slow other routers. This is also not new. However, the appropriate point of reference is "RFC 1266 - Experience with the BGP Protocol" (http://www.faqs.org/rfcs/rfc1266.html). Read section 9 -- this points to how packet loss results in BGP failures and points to how ensuring BGP packets have priority fixes this. This was published in 1991 :-) and is generally well known.
Similarly, I haven't read the referenced NDSS paper (http://www-users.cs.umn.edu/~hopper/lci-ndss.pdf) but I am also surprised that BGP holddown timers don't prevent some of the related route churn problems.
1) Blaming corporations is the wrong path. Corporations are forced to compete on a global playing field. Our world has evolved to the point that having large research labs (e.g., IBM TJ Watson, AT&T Research, Bell Labs, MSR) is not an effective competitive advantage. The problem is that, in many fields of research, it is too easy to play catch-up. These research labs put out a few great ideas that are brought to market (but the labs cost a large number of $). These ideas are copied (i.e. people come up with similar techniques that circumvent patents) sufficiently well in a few months. As a result, the profit gained by being first to market is limited and the business of creating ideas is no longer profitable for individual companies :-(
Unfortunately, we still need research to advance and grow entire markets - the problem is identifying a way to fund or compensate companies/individuals/organizations for their research contributions without relying on market profits.
Basically, the world changed - rapid prototyping and easy access to ideas/research are common to most topic areas. It's time to face the facts and come up with something new that works.
2) Foreigners and graduate school. This is a huge problem. The American graduate education system has survived on the influx of students from countries like India and China. Unfortunately, the combination of improving opportunities in these countries and a negative immigration environment in the US has made this transfer of talent slow down. In 20 years, I am not convinced that the best and brightest will continue to come to the U.S.
3) No jobs for PhDs. I agree that jobs are getting harder to find for PhDs. Given comment (1) above, I am not sure there are going to be more PhD research jobs in the future. If we solve (1), there will be more jobs.
This looks a lot like Pioneer:
Seshadri, Arvind, Mark Luk, Elaine Shi, Adrian Perrig, Leendert van Doorn, and Pradeep Khosla.
"Pioneer: Verifying Integrity and Guaranteeing Execution of Code on Legacy Platforms."
In Proceedings of the ACM Symposium on Operating Systems Principles (SOSP), Brighton, United Kingdom, October 2005.
http://sparrow.ece.cmu.edu/~adrian/projects/pioneer.pdf